Using Linked Claims and EHR Real-World Data to Determine Medication Adherence Among Common Therapeutic Areas of Chronic Disease

OBJECTIVES: Medication adherence is vital to disease management for patients with chronic illness. This study examined medication adherence among adult patients with selected common chronic conditions, using linked commercial claims and EHR data from 2015-2023.

METHODS: De-identified Oracle electric health record (EHR) data for ~110M patients over 139 health organizations across the US was used to identify patients aged 18-64 who were diagnosed with type 2 diabetes, asthma, migraine, anxiety disorder, or insomnia. Analysis included patients who had at least one outpatient encounter between January 1, 2015, and June 30, 2023, and had a medication prescribed at the encounter. Pharmacy claims from a nationally representative US closed claims database for ~90M EHR-linked patients from the same time period were used to determine whether prescriptions were filled. Prescriptions were considered filled if they were filled within 30 days after the encounter. Patients who filled 80% or more of their prescriptions were considered adherent. Patient-level drug adherence was compared across therapeutic areas and drug-level fill rates were compared across drug classes.

RESULTS: Among 95,980 linked diabetes patients, 45.3% were adherent. Out of 606,653 diabetes medication records, amylin analogs (73.7%) and GLP-1 receptor agonists (63.8%) had the highest fill rates, while antidiabetic combination medications (52.4%) had the lowest fill rate. Among 33,966 linked asthma patients, 45.6% were adherent. Out of 119,136 asthma medication records, leukotriene modifiers (64.8%) and methylxanthines (61.0%) had the highest fill rates, while adrenergic bronchodilators (53.3%) had the lowest fill rate. Results for migraine, anxiety disorder, and insomnia will be discussed in the poster.

CONCLUSIONS: Using EHR and claims data from nationally representative databases, low medication adherence rates were observed among linked patients with diabetes and asthma, where slightly less than half were found to have filled their prescribed medication at least 80% of the time, though adherence rates vary by medication.